Introduction

Using the heartData data set provided by the University of Sydney, i explored the correlation between BMI, Systolic blood pressure, Diastolic blood pressure and death. I initially theorised that a high BMI would correspond directly with mortality, as well as show a higher degree of correlation with diastolic blood pressure than with systolic blood pressure

Results

#diastolic bp vs BMI
ggplot(heartData, aes(BMI, DIABP) ) +
  geom_point( aes(color = DEATH) )

fig1

#""density map
ggplot(heartData, aes(x=BMI, y=DIABP) ) +
  geom_bin2d(bins = 70) +
  scale_fill_continuous(type = "viridis") +
  theme_bw()

fig2

Figure 1 displays the correlation between BMI, diastolic blood pressure, and death. The figure demonstrates a pronounced correlation between Diastolic blood pressure and BMI, though figure 2, which displays the density of data points, reveals that this correlation is not especially pronounced. It is difficult to confirm definitively that a higher diastolic blood pressure and BMI increases the likelihood of mortality due to clustering and overlapping data points, however it does appear that there is a higher density of deceased individuals with high diastolic blood pressure. Outliers from the cluster also appear to be more likely to be deceased.

fig3 fig4

Figure 3 displays the correlation between BMI and systolic blood pressure. there does appear to be somewhat of a correlation between BMI and systolic blood pressure, Figure 4 shows that though variation may be greater, there is a stronger correlation between BMI and systolic blood pressure, when observing the density of points. There also appears to be a much stronger correlation between mortality and systolic BP, compared to diastolic BP. This is again illustrated by the clustering of light blue data points around the top of the range, where as dark points, indicating surviving participants, clusters around the bottom of the normal range.

#systolic vs diastolic
ggplot(heartData, aes(DIABP, SYSBP) ) +
  geom_point( aes(color = DEATH) )

fig5

ggcorr(heartData, method = c("everything", "pearson"))

fig6

Figure 5 simply confirms the hypothesis that diastolic blood pressure is correlated to systolic blood pressure. It also appears that high overall blood pressure is more likely to lead to death, with almost all outliers also succumbing within the time course of heartData study. This is substanciated by figure 6, a corrilation chart which displays stronger corrilation between 2 variables as darker red, the corrilation between systolic BP and death is clearly stronger than tha between diastolic BP and death

Conclusion

The relationship between systolic / diastolic blood pressure, BMI and death is quite complex, yet it is apparent that an increased BMI is corilated with increased blood pressure, which in turn increases the likelyhood of death